I'm Julius, a robotics and machine learning engineer in San Francisco.
I’m a software engineer at Nimble Robotics. My work there spans reinforcement learning, Python-to-Rust policy deployment, and inference optimization. My research background is in robot learning and sim-to-real transfer.
Selected work
- Dynamic cloth folding — learning visual feedback policies and transferring them to a physical robot. First-author IROS 2022 paper; Best Paper Award finalist.
- Robot evaluation — contributions to the ICRA 2024 cloth competition’s evaluation tooling. Co-author of the resulting IJRR benchmark paper.
Merged upstream contributions
- PyTorch ExecuTorch — React Native iOS LLaMA demo and native bridge.
- tch-rs — tensor-expression fuser configuration binding for Rust.


